www.nature.com/scientificreports OPEN Opportunistic migration and habitat use of the giant mottled eel Anguilla marmorata (Teleostei: Received: 27 October 2017 Accepted: 26 March 2018 Elopomorpha) Published: xx xx xxxx Takaomi Arai 1 & Naoko Chino2 Freshwater eels have fascinated biologists for centuries due to the spectacular long-distance migrations between their freshwater habitats and their spawning areas far out in the ocean. Although freshwater eels originated in the Indonesian region, remarkably little is known about the life history of tropical freshwater eels. The diverse migratory patterns and habitat choice between marine and freshwater environments by the giant mottled eel Anguilla marmorata Quoy & Gaimard, 1824 were examined by analysing the otolith strontium (Sr) and calcium (Ca) concentrations collected in Asian waters. The wide range of otolith Sr:Ca ratios indicated that the habitat use of A. marmorata was opportunistic among fresh, brackish, and marine waters. The present study frst confrmed the occurrence of marine- resident eels that have never migrated into a freshwater habitat in A. marmorata. A. marmorata may have the same behavioural plasticity as temperate and other tropical anguillid species regarding whether to enter freshwater or to remain in estuarine and marine environments. Freshwater eels migrate fexibly among freshwater, brackish water, and seawater environments and it is now evident that their movement into freshwater is not an obligate migratory pathway but should be defned as an opportunistic catadromy, with marine and estuarine residents as ecophenotypes. Freshwater eels of the genus Anguilla, being catadromous1, migrate between freshwater growth habitats and of- shore spawning areas. Nineteen species/subspecies of Anguilla have been reported world-wide, thirteen of which are known to occur in tropical regions2,3 that are globally distributed in temperate, tropical, and subtropical areas. In general, freshwater eels are divided into temperate and tropical eels, based on their major distribution and ecological properties3. Te tropical species are thought to be more closely related to the ancestral form than their temperate counterparts. Tus, studying the life history and migration of tropical eels may provide some clues to understanding: the nature of primitive forms of catadromous migration in anguillid eels; and how the migration of the species became established. Te giant mottled eel Anguilla marmorata is a unique tropical freshwater eel that reaches large sizes of 2 m in length with a maximum weight of 21 kg4. Tis species has the widest geographic distribution of the 19 species/ subspecies of freshwater eels2 and is found longitudinally from the east coast of Africa to the Marquesas Islands in the southeast Pacifc Ocean and as far north as southern Japan2. Tis species was also found at the Palmyra Atoll in the central Pacifc5 and even farther to the east in the Galapagos Islands6, which may indicate that it has an even wider geographic range than previously thought. Because of the wide geographic range of A. marmorata, which is separated by several major landmasses, it is clearly unlikely to comprise a single panmictic population such as temperate anguillid species found in one particular region of a single ocean basin7–11. Terefore, the ecological and biological characteristics such as reproduction, life history, migration and habitat use in A. marmorata seem to be diferent for each population. Information regarding the life history characteristics of the tropical eels has been gradually accumulated12–22. Hatching was estimated to occur throughout most of the year with a constant age at recruitment15. Year-round recruitment of tropical glass eels to the river mouth12,19 follows year-round spawning18,21 and a stable recruitment 1Environmental and Life Sciences Programme, Faculty of Science, Universiti Brunei Darussalam, Jalan Tungku Link, Gadong, BE, 1410, Brunei Darussalam. 2Atmosphere and Ocean Research Institute, The University of Tokyo, 5-1-5, Kashiwanoha, Kashiwa, Chiba, 277-8564, Japan. Correspondence and requests for materials should be addressed to T.A. (email: [email protected]) SCIENTIFIC REPORTS | (2018) 8:5666 | DOI:10.1038/s41598-018-24011-z 1 www.nature.com/scientificreports/ Total Length (mm) Body Weight (g) Migration pattern estimated Sampling location Sample size Mean ± SD Range Mean ± SD Range from otolith microchemistry Poso, Indonesia 15 723 ± 282 494–1290 1350 ± 128 400–3950 freshwater/estuarine/marine residents Amami Islands, Japan 51 550 ± 80.6 367–787 415 ± 241 87.2–1280 freshwater/estuarine residents Bonin Islands, Japan 15 918 ± 189 754–1350 2208 ± 1520 485–6000 freshwater/estuarine residents Quang Tri, Vietnam 12 581 ± 209 342–1030 724 ± 1010 90.4–3390 freshwater/estuarine residents Quang Ngai, Vietnam 12 516 ± 275 355–1138 728 ± 1400 100–4320 freshwater/estuarine/marine residents Binh Dinh, Vietnam 10 884 ± 159 671–1120 1990 ± 1010 980–3270 freshwater/estuarine/marine residents Table 1. Specimens of the tropical eel Anguilla marmorata collected from six localities in Indonesia, Japan and Vietnam waters used in the present study. age15. Such a life history strategy difers from that of the temperate eels, which have a limited spawning season fol- lowed by a limited period of recruitment. However, little information is available on the tropical eels regarding their life history and migration during the yellow and silver eel stages afer recruitment to coastal waters as glass eels. Te migratory history of several species of freshwater eels have been studied using microchemical techniques that determine the ratio of strontium to calcium (Sr:Ca ratio) in their otoliths. Te Sr:Ca ratio in the otoliths of freshwater eels difers according to the time they spend in fresh water versus sea water23. Studies on strontium incorporation into eel otoliths of Anguilla japonica showed that the Sr:Ca level in their otoliths strongly correlated with the salinity of the water and was little afected by other factors such as water temperature, food and physio- logical factors24. Recently, Arai and Chino25 found that the Ca and Sr contents and the resultant Sr:Ca ratios in the rearing water signifcantly increased with salinity in the giant mottled eel Anguilla marmorata. Tus, the Sr:Ca ratios of otoliths could help to determine whether individual eels actually enter fresh water at the elver stage and remain in a fresh water, estuarine or marine environment until the silver eel stage, or whether they move between diferent habitats with difering salinity regimes. Te objectives of this study were to accumulate ecological and biological information regarding the migra- tion afer recruitment to coastal waters of the giant mottled eel Anguilla marmorata collected in East Asia and Southeast Asia, as there has been little available information concerning its migration. Furthermore, we discuss the evolution and occurrence of the diverse migrations of the eels based on a comprehensive analysis of the knowledge from present and previous studies on their life history. Results Biological characteristics. Te total length (TL) of Anguilla marmorata ranged from 342 to 1350 mm with a mean ± SD of 649 ± 229 mm (n = 151). Te body weight (BW) ranged from 87 to 6000 g with a mean ± SD of 822 ± 1090 g (n = 151) (Table 1). Migratory history. In all specimens examined, the otoliths had a central region of high Sr:Ca ratios that corresponded to the leptocephalus stage and showed lower Sr:Ca ratio levels afer metamorphosis into the glass eel stage (Fig. 1). Each otolith had a peak between the otolith core and approximately 150 µm outward. Te ratios in the otoliths of eels before the elver stage were similar among specimens, indicating that the migratory history was similar among specimens during the oceanic leptocephalus stage. Outside of the high Sr:Ca core, there were strong patterns in the Sr:Ca ratios of the eels’ otoliths. Te patterns in the otolith Sr:Ca ratios from Poso River of Indonesia were classifed into three constant types (Fig. 1a,b,c): (1) constantly low values of 1.23–1.77 × 10−3 (mean: 1.54 ± 0.20) (n = 6, Fig. 1a), (2) intermediate values of 2.56–5.95 × 10−3 (mean: 4.30 ± 1.80) (n = 4, Fig. 1b), and (3) relatively high values of 6.17–7.71 × 10−3 (mean: 7.17 ± 0.59) but larger than 6 × 10−3 with no movement into freshwater (n = 5, Fig. 1c). Signifcant diferences in average Sr:Ca values between each type were found for all combinations (Kruskal-Wallis test, p < 0.05–0.0001). Te life history migratory patterns from Amami Islands in Japan were generally classifed into two types (Fig. 1d,e): (1) constantly low values of 1.44–1.98 × 10−3 (mean: 1.74 ± 0.18) (n = 10, Fig. 1d), (2) intermediate values of 2.02–3.92 × 10−3 (mean: 2.91 ± 0.52) (n = 41, Fig. 1e). Tere were signifcant diferences in average Sr:Ca values between the types (Mann-Whitney U-test, p < 0.0001). However, no eel showed relatively high Sr:Ca ratios more than 6 × 10−3 along the life history transect, which would indicate marine residence. Te eels from Bonin Islands of Japan showed unique and diverse migratory patterns classifed into a total of fve types with apparently two resident (Fig. 1f,g) and three migrant patterns (Fig. 1h,i,j). Tese two residents were (1) constantly low values of 1.51–1.92 × 10−3 (n = 2, Fig. 1f) and (2) intermediate values of 2.03–3.41 × 10−3 (mean: 2.48 ± 0.50) (n = 7, Fig. 1g). Tere were three groups of migrant types, with Sr:Ca ratios shifing between the inner and outer portion of the transect (Fig. 1h,i,j). In the frst group (3), Sr:Ca ratios were low in the inner portion were low (range: 1.71–1.91 × 10−3; mean ± SD: 1.79 ± 0.10) and then increased to a middle range (range: 3.50–5.48 × 10−3; mean ± SD: 4.79 ± 1.12) (Fig.